A stereo camera calibration based on robotic vision

Abstract

Robotic vision has become a very popular field in recent years due to the numerous promising applications it may enhance. However, errors within the cameras and in their perception of their environment can cause applications in robotics to fail. To help correct these intrinsic and extrinsic imperfections, stereo camera calibrations are performed. There are currently many accurate methods of camera calibration available; however, most or all of them are barely theoretical and difficultly practical. Because our robot need to accurately approach and placement scientific objects, in this paper we present a image rectification method that is an extension to two-step method. There is an additional step for correcting the distorted image coordinates. The image rectification is performed and accurately compensates for radial and tangential distortions. Finally, through Matlab tool developed the results of experiment are accurate and available.